Discovery

What is discovery?

It starts with questions

Data exploration

Data analysis

Discovery tools

Spreadsheets (e.g. Excel, OpenOffice)

Tableau

R + ggplot2*

Python / ipython / pandas / matplotlib*

(* = free)

Tableau

R + ggplot2

  qplot(Sepal.Length, Petal.Length, data = iris, color = Species,
	  size = Petal.Width)
						

blog.echen.me/2012/01/17/quick-introduction-to-ggplot2/

iPython Notebook / matplotlib

titanic = pd.read_excel("data/titanic.xls", "titanic")
bp = titanic.boxplot(column='age', by='pclass', grid=False)
for i in [1,2,3]:
    y = titanic.age[titanic.pclass==i].dropna()
    x = np.random.normal(i, 0.04, size=len(y))
    plot(x, y, 'r.', alpha=0.2)

Search for fonnesbeck plotting and visualization

Presentation

Desktop

Browser-based

Desktop

Excel

Tableau

R + ggplot2 (free)

Python / ipython / pandas / matplotlib (free)

NodeBox (free)

Excel

http://www.datadrivenconsulting.com/2010/02/data-visualization-challenge-my-dashboard-design/

Excel

Power Map plug-in

Tableau examples

See http://www.tableausoftware.com/public/gallery

R / ggplot2

Python / matplotlib

Python / bokeh

Example

NodeBox 1

NodeBox

NodeBox 3

NodeBox

Browser-based
(JavaScript libraries)

Highcharts, Google Charts etc.

D3

Custom coded (SVG, Canvas)

See also datavisualization.ch

Highcharts, Google Charts etc.

Off-the-shelf charts

D3

Usually requires custom coding

d3js.org

Built with D3

Custom coded

Driven from tab-separated-value data

www.nytimes.com/interactive/2013/05/25/sunday-review/corporate-taxes.html

My approach

Primarily programmatic (JavaScript)

More often than not, driven from spreadsheet (CSV) data

Usually use D3, but not always

Discovery

Mix of Excel & custom scripts

Moving towards Python, pandas etc.

Presentation

Browser based

Load data in from csv or JSON

Simple reformatting

D3 & custom code to present data

Example

Data from World Bank

Matrix of migratory flows (spreadsheet)

Convert into JSON format using Python (offline)

{"Canada": 
  {"totalIn": 7202340,
  "totalOut": 1185392,
  "out": [
    {"country": "United States", "amount": 834945}
  ],
  "in": [
    {"country": "China", "amount": 543573},
    {"country": "Germany", "amount": 199541},
    {"country": "Guyana", "amount": 101505},
    {"country": "Hong Kong, China", "amount": 250786},
    etc.
						

Python is often used server-side so this could be live data...

Simple webpage (one HTML file)

Load in JSON using JavaScript

Draw map using SVG, JavaScript and D3